Re: [Banshee-List] Question about MusucBrainx Metadata

[Julien Sobrier <julien sobrier net>]

Aaron Bockover wrote:
> This wouldn't actually change the amount of data downloaded.
> XmlTextReader is used on the HTTP stream, so the XML is processed in
> "real time" as the stream is read. When enough data has been read, the
> stream is closed. It's super fast. Basically, the XML document never
> gets fully downloaded or parsed into a DOM tree, so the IO and memory
> overhead is extremely minimal.
> 
> --Aaron

Thank you very much for the explanation.